LDCad 1.4a (win+linux)


LDCad 1.4a (win+linux)
#1
I just released LDCad 1.4a it fixes a couple of problems and includes new snap info and part bin tweaks.

Like always you'll find the newest version at:

http://www.melkert.net/LDCad/download

Biggest change is the new 'figures' branch inside the sorted part bin group. It allows for easy access to the new friends figures, it also includes the old minifig groups (which I tweaked a bit in order to incorporate the new arms) and new Fabland and Technic figure groups.

I've also added part snapping to all parts in those groups except for the minifig hair/hat and accessories.

And I tried to apply some of the smaller requests from the 1.4 thread, like: ESC to deselect and allowing lower FOV etc.

Finally as LDCad now is included in the AIOI I also updated my own installer to look for the AIOI location of LDCad in order to apply updates as the AIOI installs in "Program Files (x86)\LDraw\LDCad" instead of "Program Files (x86)\LDCad" by default.


With this version released I'm back to working on 1.5, which will probably take another couple of months to complete. But like always don't hesitate to report problems / feature requests.
Reply
Re: LDCad 1.4a (win+linux)
#2
Right on, going to install it asap. Thanks for your hard work.
______________________________________________
OS = Ubuntu 14.04 LTS (64bit)
Reply
Re: LDCad 1.4a (win+linux)
#3
Excellent!
Quote:Biggest change is the new 'figures' branch inside the sorted part bin group. It allows for easy access to the new friends figures, it also includes the old minifig groups (which I tweaked a bit in order to incorporate the new arms) and new Fabland and Technic figure groups.
Highly appreciated, after all the time I spent on Fabuland and Friends figures, to be able to edit them easily!
Quote:I've also added part snapping to all parts in those groups except for the minifig hair/hat and accessories.
Well, the Friends hairs I tested do have accessory pin connectivity?

Speaking of connectivity, it seems you didn't update connectivity on smooth panels and 1x1 round tiles?
(edit) Is it intended that there is no stud connectivity in the middle of stud4 primitives?

Parts bins:
- I like a lot the new figure bin organization. A tiny glitch: "Figure Fabuland hibs and limb" instead of "Figure Fabuland hips and limb"
- I think that mixing pins and connectors in the same Technic bin is not such a good idea, as there are MANY connectors, and pins are so ubiquitous. Pins are not very easy to use, especially since many connectors have pin in description making filter uneffective.
Reply
Re: LDCad 1.4a (win+linux)
#4
Quote:Highly appreciated, after all the time I spent on Fabuland and Friends figures
Right back at you Smile

Quote:Well, the Friends hairs I tested do have accessory pin connectivity?
I mean the normal minifig aso/hair/heads as there are way more of those.

Quote:....smooth panels and 1x1 round tiles
I thought I fixed those, I checked again and noticed the tech pin not wanting to go to both sides of some of the holes. That's actually caused by the pin it self as axles etc do behave correctly. Anyhow I corrected it.

As for the plates (6141) do you mean to place them stud facing outwards inside a tech beam hole, like so:

.png   1x1.png (Size: 20.37 KB / Downloads: 4)
If so I've added that info to the 1x1 plates. I didn't notice anything else wrong with the 1x1 tiles though.

Quote:(edit) Is it intended that there is no stud connectivity in the middle of stud4 primitives?
I'm not completely sure but I think I left stud4.dat without snap info because I considered it not a 'real' stud primitive. This in means it is used scaled and the hole will almost always lead into some bigger shape in the higher parts.

I'll have to do some more looking around and try to make the needed adjustments for 1.5.

Quote: I think that mixing pins and connectors in the same Technic bin is not such a good idea, as there are MANY connectors, and pins are so ubiquitous. Pins are not very easy to use, especially since many connectors have pin in description making filter uneffective.
Yes, most bins are based upon a set of include/exclude filter rules and it is there fore very hard to split the connectors from the pin parts unless it's done by handpicking the .dat numbers which isn't very future proof. Do you know of another way to uniformly split those parts based on description, number, keyword, etc?


I redid the release packages of 1.4a (as the executable's remain the same) and updated the downloads. Alternativly you could place the attached changed seed files in your existing setup (rename to .sf)


Attached Files
.zip   partBin.zip (Size: 431.24 KB / Downloads: 0)
.zip   shadow.zip (Size: 290.43 KB / Downloads: 0)
Reply
Re: LDCad 1.4a (win+linux)
#5
Quote:I thought I fixed those, I checked again and noticed the tech pin not wanting to go to both sides of some of the holes. That's actually caused by the pin it self as axles etc do behave correctly. Anyhow I corrected it.
Yes, that was that. Works great now (on a fresh reinstall of 1.4a+ on another machine).
Quote:As for the plates (6141) do you mean to place them stud facing outwards inside a tech beam hole, like so:
No, I meant plain 6141 tube to any regular stud connexion. No snapping here. 6141p01 did work though so I had a look in shadow.sf: 6141 was missing. I created one from 6141p01 connection file and now it works.
Quote:I'm not completely sure but I think I left stud4.dat without snap info because I considered it not a 'real' stud primitive. This in means it is used scaled and the hole will almost always lead into some bigger shape in the higher parts.
I'll have to do some more looking around and try to make the needed adjustments for 1.5.
I see your point, yes stud4 is often scaled. That wouldn't be such a big issue if only its origin was placed on the rim...
Quote:Yes, most bins are based upon a set of include/exclude filter rules and it is there fore very hard to split the connectors from the pin parts unless it's done by handpicking the .dat numbers which isn't very future proof. Do you know of another way to uniformly split those parts based on description, number, keyword, etc?
I see your point here too Wink No simple solution unfortunately. Do you have a short description of filter syntax? (especially, what does "exclude !description pin,connector,bush" do?)
Reply
Re: LDCad 1.4a (win+linux)
#6
> That wouldn't be such a big issue if only its origin was placed on the rim...

maybe we can create a new primitive which has that property
and make the old one a MOVED TO that?
Reply
Re: LDCad 1.4a (win+linux)
#7
Philippe Hurbain Wrote:6141p01 did work though so I had a look in shadow.sf: 6141 was missing. I created one from 6141p01 connection file and now it works.
Very weird, 6141 is just an alias of 4073 which does have info. 6141p01 has it's own info as it's not based upon a plain 6141. anyway there is something fishy going on as I seem to get normal snapping for 6141 in a clean install myself. I'll have to investigate this some more.

Philippe Hurbain Wrote:Do you have a short description of filter syntax? (especially, what does "exclude !description pin,connector,bush" do?)

The list of include / exclude lines in pbg files of the filter kind build up a list by ether adding parts to the current result or removing some from the result based upon the secondary rule.

The second item in a rule is the search field, it can be one of the following

description
category
keyword
partName

When prefixed with '!' the rule inverses like: 'must not match'.

The final part of the rule is a list of phrases the selected field should match. You can list multiple ones by separating them using ','. By default they will act as "*text*" wildcards, but you can also use e.g "text*" to match anything starting with the phrase.

So keeping this in mind your example "exclude !description pin,connector,bush" means:

Remove all items not having the text: pin, connection or bush in their description from the current result list.

Hope this helps.
Reply
Re: LDCad 1.4a (win+linux)
#8
> Very weird, 6141 is just an alias of 4073

note that these files have swapped roles now.
the PT has a different MOVED-TO than the previously official version.

The instances on the PT are a fix for the previous situation.
Reply
Re: LDCad 1.4a (win+linux)
#9
That's probably what's playing here, if Philio is using the full unofficial library it will 'suppress' the 1402 situation resulting in said problems.

Thanks for pointing this out Steffen I'll make a note of it so I can correct it once 1501 comes out.
Reply
Re: LDCad 1.4a (win+linux)
#10
Ah - that was _THE_ reason... thanks, Steffen & Roland!
Reply
Re: LDCad 1.4a (win+linux)
#11
Quote:Hope this helps.
It does. The following bins suit my taste:
Code:
[options]
kind=basic
caption=Technic
description=Technic parts
mascot=4019.dat

<items>
sorted\technicAxles
sorted\technicBeams
sorted\technicBricks
sorted\technicPins
sorted\technicCons
sorted\technicGears
sorted\technicPanels
sorted\technicPlates
sorted\technicPneumatic
sorted\technicMotors
sorted\technicMindStorm
sorted\technicShocks
sorted\technicSteering
sorted\technicBionicle
sorted\technicMisc
Code:
[options]
kind=filter
caption=Pins
description=Technic pins
mascot=3673.dat
sortOn=description


<rules>
include category technic
exclude description ~*
exclude description _*
exclude description =*
exclude description brick,plate,beam,bionicle,rubber band holder,steering,blade,panel,gear
exclude !description Technic pin, Technic axle pin, Technic Bush
Code:
[options]
kind=filter
caption=Connectors
description=Technic connectors
mascot=6536.dat
sortOn=description


<rules>
include category technic
exclude description ~*
exclude description _*
exclude description =*
exclude description brick,plate,beam,bionicle,rubber band holder,steering,blade,panel,gear,Technic pin, Technic axle pin, Technic bush
exclude !description pin,connector,bush
Reply
Re: LDCad 1.4a (win+linux)
#12
Configuration files are fun aren't they Smile

I do like the idea of splitting the connectors and pin parts into separate bins, so I might add it to the 1.5 default bin.

But keep in mind these modified files might get overwritten during updates unless you setup a whole second (copy) part bin besides the 'default' folder one.
Reply
Re: LDCad 1.4a (win+linux)
#14
Roland Melkert Wrote:But keep in mind these modified files might get overwritten during updates unless you setup a whole second (copy) part bin besides the 'default' folder one.

I mess with the main.gui file and sort out any changes every update to keep my colors and now I'll add this to mix. No biggie though, a few minutes of tweaking to make it the way you like it is very worth it.
______________________________________________
OS = Ubuntu 14.04 LTS (64bit)
Reply
Re: LDCad 1.4a (win+linux)
#16
One drawback: part 55615 (Technic Pin Connector perpendicular etc...) is sorted with pins instead of connectors. I have not yet found a solution (except renaming part...)
Reply
Re: LDCad 1.4a (win+linux)
#17
Philippe Hurbain Wrote:One drawback: part 55615 (Technic Pin Connector perpendicular etc...) is sorted with pins instead of connectors. I have not yet found a solution (except renaming part...)

You could use

Code:
keep partName 55615.dat
in the connectors group, add it before the rule which otherwise would cause the removal (after the =* excl).

And if you don't want it in the pins group use
Code:
exclude partName 55615.dat
as the last line in that one.


I didn't describe the 'keep' rule as it's a bit messy for the exact same reason you ran into. In short the rule assigns exclude immunity to any items it matches in the current result set.

There's also the 'static' rule, which is even more messy it will add anything to the current set and also assigns it immunity to any following rules. I don't recommend using ether rule unless absolutely necessary Smile
Reply
Re: LDCad 1.4a (win+linux)
#18
Perfect!
Reply
Re: LDCad 1.4a (win+linux)
#20
+1 to that, I just wished that today while I was messing about with a model. For some reason I was thinking I could so I tried and ended up flipping my model upside-down lol.
______________________________________________
OS = Ubuntu 14.04 LTS (64bit)
Reply
Re: LDCad 1.4a (win+linux)
#13
Philippe Hurbain Wrote:It does. The following bins suit my taste:

Oh, I love that. Technic is about all I work with so that is awesome.
______________________________________________
OS = Ubuntu 14.04 LTS (64bit)
Reply
Please note (was LDCad 1.4a..)
#15
Hi all,

It just came to my attention there is a very severe bug in the new 1.4a version as a result of one of the earlier bugfixes.

As a result there is a very high chance on crashing the program while adding a submodel to a mpd if that model also holds one or more bendable parts (path content).

I will release a 1.4b later tonight or tomorrow as this bug disrupts core usage, in the mean time it's advised to do a ctrl+s before adding submodels and / or make sure the path content is never rendered again afterwards.

Alternatively you could also temporary use the old 1.4 version executable while keeping all the other 1.4a files.

Sorry for the inconvenience Wink
Reply
Re: LDCad 1.4a (win+linux)
#19
Quote:But like always don't hesitate to report problems / feature requests.
One of the things I really miss is some form of "mass selection", a "select everything behind the rectangle" tool would be a very good start...
Reply
Re: LDCad 1.4a (win+linux)
#23
Philippe Hurbain Wrote:One of the things I really miss is some form of "mass selection", a "select everything behind the rectangle" tool would be a very good start...
This is pending, if time permits it will be in the 1.5 version.
Reply
Re: LDCad 1.4a (win+linux)
#24
Another idea, for what it is worth (dunno if it fits LDCad architecture): placing a large submodel in a large model can be tricky using parts snapping, so many possible connections... SR3D only tries to fit the currently selected part, drastically reducing the number of possible connections, speeding up things. This implies that the part used to hold submodel somehow becomes that key part.

(edit) Maybe it would be a good idea that moving the stiffness control in flex parts used the grid. This would ease realisation of symmetrical flex.
Reply
Re: LDCad 1.4a (win+linux)
#25
Philippe Hurbain Wrote:Another idea, for what it is worth (dunno if it fits LDCad architecture): placing a large submodel in a large model can be tricky using parts snapping, so many possible connections... SR3D only tries to fit the currently selected part, drastically reducing the number of possible connections, speeding up things. This implies that the part used to hold submodel somehow becomes that key part.
Yes this bothers me too, it's the reason you can disable snapping at anytime using shift+p. I do like the snap to selection only approach but I don't think I want it all the time. I'll see if I can set something up using ether options or some smart rules on when to use it.

The long term solution is also on my to do list which is excluding 'occupied' connections so things snap a whole lot more useful while using multiple parts / submodels and some of the larger parts.


Philippe Hurbain Wrote:(edit) Maybe it would be a good idea that moving the stiffness control in flex parts used the grid. This would ease realisation of symmetrical flex.
I like this, I'll make it an option and/or I'll add an extra field to the grid settings.

Thanks for the input.
Reply
Re: LDCad 1.4a (win+linux)
#26
I think that LDCad should have a standard angular grid at 22.5° (the step of click hinges and Technic angle connectors). More annoying, I can't create a new one because I can't enter decimal angles (OK, 22° made a fairly good approximation Wink. What is the Freq parameter in grid stepping configuration?
Reply
Re: LDCad 1.4a (win+linux)
#27
Philippe Hurbain Wrote:I think that LDCad should have a standard angular grid at 22.5° (the step of click hinges and Technic angle connectors). More annoying, I can't create a new one because I can't enter decimal angles (OK, 22° made a fairly good approximation Wink.
At the time I was thinking fractions are only needed when using the dialog. But you are right fractions (or at least .5) should be useable, I'll put it on my todo list.

Philippe Hurbain Wrote:What is the Freq parameter in grid stepping configuration?
It's used for the rendering of the grid lines / rotation disc segments. With move it governs the frequency of the the white lines vs the grey ones, with rotation it controls how many steps a segment on the disc represents.
Reply
Re: LDCad 1.4a (win+linux)
#21
For the record, connectivity of pegholes of 11949 seem to be off by 2 ldu...
Reply
Re: LDCad 1.4a (win+linux)
#22
Thanks, fixed in the attached seed file.

Like always rename to .sf and place in the seeds folder while the program is closed.


Attached Files
.zip   shadow.zip (Size: 290.58 KB / Downloads: 0)
Reply
Re: LDCad 1.4a (win+linux)
#28
Roland Melkert Wrote:With this version released I'm back to working on 1.5, which will probably take another couple of months to complete. But like always don't hesitate to report problems / feature requests.

What I'm really missing right now is a Minifig Generator.

w.
LEGO ergo sum
Reply
Re: LDCad 1.4a (win+linux)
#29
Well, with parts snapping and rotation along local part axes, it's nearly useless. Things useful to know:
- switch to part local axes: select part, then press 'o'. Switch back to world axes: press 'o' with nothing selected. Eg. to rotate arm/hand/accessory, select them all (with arm selected first), press 'o' then rotate everything. Similar to rotate hand+accessory.
- To use parts that don't have yet connectivity: use a similar part that already has connectivity defined, then select it and double click in parts bin on the right part to replace the place holder.
Reply
Re: LDCad 1.4a (win+linux)
#30
Philippe Hurbain Wrote:Well, with parts snapping and rotation along local part axes, it's nearly useless.

It's still more time consuming to stitch a figure together selecting parts from the library then selecting from a drop-down (where a small preview would not only show the entire figure, but also the selected part when I hover over the drop-down entries just to get additional information to the textual description).

Philippe Hurbain Wrote:press 'o' with nothing selected.

Since you talk about: I still find it tiresome to unselect via STRG+SHIFT+a. A single click somewhere on the empty workbench, (say where there are no parts) would be IMHO much more intuitive.

w.
LEGO ergo sum
Reply
Re: LDCad 1.4a (win+linux)
#31
Willy Tschager Wrote:It's still more time consuming to stitch a figure together selecting parts from the library then selecting from a drop-down (where a small preview would not only show the entire figure, but also the selected part when I hover over the drop-down entries just to get additional information to the textual description).
Yeah... and I must admit that many (most?) accessories are still missing a bar connectivity (and minifig hand is missing a stud to tube connectivity on top).

Quote:Since you talk about: I still find it tiresome to unselect via STRG+SHIFT+a. A single click somewhere on the empty workbench, (say where there are no parts) would be IMHO much more intuitive.
I mostly agree with this one... though it is sometimes useful to avoid unintended deselection after a long multiple selection session!
Reply
Re: LDCad 1.4a (win+linux)
#32
Willy Tschager Wrote:It's still more time consuming to stitch a figure together selecting parts from the library then selecting from a drop-down (where a small preview would not only show the entire figure, but also the selected part when I hover over the drop-down entries just to get additional information to the textual description).
I have been thinking about a minifig (or generic figure) generator but couldn't decide on how to integrate it. With part snapping in 1.3 I, like Philippe, thought it wouldn't be necessary anymore.

That said I do have some new ideas on making figure editing easier, so something extra will pop up sooner or later.

Quote:Yeah... and I must admit that many (most?) accessories are still missing a bar connectivity (and minifig hand is missing a stud to tube connectivity on top).
I'll add the top studs and try to complete all aso parts soon.

Quote:Since you talk about: I still find it tiresome to unselect via STRG+SHIFT+a. A single click somewhere on the empty workbench, (say where there are no parts) would be IMHO much more intuitive.

As of 1.4a you can also deselect using the esc key. Someone, I think it was Philippe Smile, suggested this in one of the older release threads.
Reply
Re: LDCad 1.4a (win+linux)
#33
Oops, indeed Wink
Reply
Re: LDCad 1.4a (win+linux)
#34
Quote:I'll add the top studs and try to complete all aso parts soon.
In the meantime (I guess this is a huge work) maybe it would be interesting to provide a hint of the level of connectivity a part offers? Something like a dot in parts bin, red = no connectivity, orange = some inherited connectivity, green = full connectivity. Of course this is only interesting if it can be generated automagically, otherwise if a manual intervention is to be done it would probably be better using that time to define connectivity Wink
Reply
Re: LDCad 1.4a (win+linux)
#35
I've been thinking about something similar as it's sometimes hard to update snapping when new LDraw libraries are released (this is also mainly why I added searching on the history lines to the part bin filter).

Maybe something like keeping track of the files I manually inspected (through a cfg file in de shadow lib) and deemed 'snap complete', those could then have some indication in the bin unless the part's history date is newer then the cfg file (so I'm forced to recheck them).

I like the dots idea to also indicated partial inherited info (without certification etc). I'll try to add it to 1.5.
Reply
« Next Oldest | Next Newest »



Forum Jump:


Users browsing this thread: 1 Guest(s)